tidyverse: data manipulation (tasks)

Biomedical Data Sciences | LUMC

dplyr : data manipulation

Load tidyverse:

library( tidyverse )

General properties of survey data

• Load the survey data in tibble format and inspect the meta information such as dimensions and column (variable) types.

• Inspect survey data with functions such as class, str, glimpse (tidyverse) functions.
  • What is the difference between str and glimpse output?

• Produce the following values on survey data by using the pipe operator %>%:

  • number of rows
  • number of columns
  • last 8 rows
  • first 7 rows
  • last 3 column names

Selection & Filtering

Apply the following to survey data:

• Rename the m.i variable into system

• Reorder the variables as such that name,age and gender come first

• Select the last three variables

• Deselect variables that relate to hand and/or arm (e.g. *.hnd, etc.)

Summary values survey data:

• Retrieve distinct values for smoking habit (smokes), do the same for exercise pattern exercise.

• Derive the frequency (count) table of smoking, do the same for exercise pattern.

• Derive the frequency (count) table of smoking and exercise pattern.

• How many females are there who never smoked?

• How many right-handed heavy smokers are there, counts per gender?

• Select teenagers.

• What are the counts of smoking habits (smokes) in teenagers?

• What are the counts of exercise patterns (exercise) in teenagers?

Add variables

• Add a new column feet with heights reported in feet unit (1 foot = 30.48 cm).

• Add a new column ‘diffHandSpan’ : the absolute difference in span of writing hand span1 and non-writing hand span2.

• Count the number of students with smaller writing hand span.

Calculate summaries and grouping

Summary (summarise)

In survey data summarise on:

• mean age along with total count

• mean writing and non-writing hand span (span1,span2) .

• mean, minimum and maximum feet (height)

Group (group_by)

1. Summarise on average age separately for each of the following group(s):

• gender

• smoking habit

• gender and smoking habit

• exercise pattern

• gender and exercise pattern

2. Repeat the previous exercise with the addition of a new column with group sizes.

Sort (arrange)

Order the survey data by:

• name

• gender and name

• name and gender

• gender,name and smoking habit

• descending order of gender, name and descending order of height

Extra exercises

1. Calculate the following in the survey data:

• average height

• minimum and maximum writing hand span span1

• minimum and maximum non-writing hand span span2

2. In the survey data some students reported their heights in metric and some in imperial units and afterwards these were converted to metric (cm). Add a new column reportedHeight with heights reported by the student in the original unit (system or m.i). Assume inch in case of imperial (1 inch = 2.54 cm) . Hint: use ifelse to test for the system used.

Copyright © 2020 Biomedical Data Sciences | LUMC